The latency-associated nuclear antigen interacts with MeCP2 and nucleosomes through separate domains

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV)-infected cells express the latency-associated nuclear antigen (LANA) involved in the regulation of host and viral gene expression and maintenance of the KSHV latent episome. Performance of these diverse functions involves a 7-amino-acid chromatin-binding motif (CBM) situated at the amino terminus of LANA that is capable of binding directly to nucleosomes. LANA interacts with additional chromatin components, including methyl-CpG-binding protein 2 (MeCP2). Here, we show that the carboxy-terminal DNA-binding/dimerization domain of LANA provides the principal interaction with MeCP2 but that this association is modulated by the CBM. Both domains are required for LANA to colocalize with MeCP2 at chromocenters, regions of extensive pericentric heterochromatin that can be imaged by fluorescence microscopy. Within MeCP2, the methyl-CpG-binding domain (MBD) is the primary determinant for chromatin localization and acts together with the adjacent repression domains (the transcription repression domain [TRD] and the corepressor-interacting domain [CRID]) to redirect LANA to chromocenters. MeCP2 facilitates repression by LANA bound to the KSHV terminal repeats, a function that requires the MeCP2 C terminus in addition to the MBD and CRID/TRD. LANA and MeCP2 can also cooperate to stimulate transcription of the human E2F1 promoter, which lacks a LANA DNA-binding sequence, but this function requires both the N and C termini of LANA. The ability of LANA to establish multivalent interactions with histones and chromatin-binding proteins such as MeCP2 would enable LANA to direct regulatory complexes to specific chromosomal sites and thereby achieve stable reprogramming of cellular gene expression in latently infected cells.

FIG. 1.

MeCP2 interacts with the C-terminal domain of LANA. (A) Domain structure of KSHV LANA, illustrating the boundaries of engineered fragments used in this study. The amino acid numbering corresponds to the prototype sequence from the BC-1 cell line (57). Numbers in parentheses signify the actual coordinates of the LANA variant used here (GenBank accession no. AAB626557) (48). (B) Results from a coimmunoprecipitation assay using T7 epitope-tagged LANA_N, LANA_NΔ22, and LANA_C fragments and Flag-MeCP2_FL. T7-HCF-1_N serves as a positive control for association with Flag-HCF-1_C (70). Human 293T cells were cotransfected with combinations of plasmids expressing the indicated constructs. After 24 h, protein extracts were prepared and immunoprecipitated (IP) using anti-Flag-coupled beads, resolved by SDS-6.5% PAGE, and immunoblotted with either anti-T7 antibody (α-T7) (top panels) or anti-Flag antibody (bottom panel). (C) The LANA CBM interacts with nucleosomes. Partially purified HeLa nucleosomes (lane 4) were incubated with GST fusion proteins corresponding to LANA residues 2 to 50 (lane 1), 23 to 50 (lane 2), and 2 to 31 (lane 3). The sequence of each LANA peptide is shown with the CBM highlighted. After extensive washing, bound complexes were heat denatured in SDS sample buffer, resolved by SDS-15% PAGE, and stained with Coomassie brilliant blue. Bands corresponding to each of the core histones are indicated.

FIG. 2.

LANA colocalizes with MeCP2 in NIH 3T3 cells. (A) NIH 3T3 cells were transiently transfected with expression plasmids encoding GFP fused to full-length LANA (GFP-LANA_FL) (61) (a, b, e, and f) and DsRed fused to full-length human MeCP2 (DsRed-MeCP2_FL) (c to j). GFP alone (g and h) and GFP-Brd2 (i and j) serve as controls. Each image panel shows a single cell and is representative of the predominant pattern in that transfection experiment. (B) MeCP2-dependent relocalization of LANA is not mirrored by histones. NIH 3T3 cells were transfected to express cytomegalovirus lacZ, DsRed-MeCP2_FL, and/or GFP-LANA_FL. After 24 h, cells were fixed and probed with an antibody against total histone H2B (α-H2B).

FIG. 3.

Colocalization with MeCP2 requires multiple regions of LANA. (A) NIH 3T3 cells were transfected with plasmids encoding DsRed-MeCP2_FL and the following GFP-LANA derivatives: GFP-LANA_FL (a and b), GFP-LANA_FLΔ22 (c and d), GFP-LANA_N (e and f), GFP-LANA_NΔ22 (g and h), and GFP-LANA_C (i and j). (B) Fusion of the CBM to LANA_C is not sufficient for colocalization with MeCP2. Cells were transfected to express GFP-LANA_CBM+C on its own (a and b) or with DsRed-MeCP2_FL (c and d).

FIG. 4.

Localization of MeCP2 to chromocenters is specified by multiple domains. (A) Domain structures of human MeCP2 and the truncation proteins used in this study. (B) The MBD is required for localization of MeCP2 to chromocenters. Representative NIH 3T3 cells expressing each of the DsRed-MeCP2 derivatives are shown. The abilities of MeCP2 derivatives to localize (+) the DsRed signal to chromocenters (foci) are summarized in panel A. (C) Immunoblot analyzed using anti-DsRed antibody to detect expression of DsRed-MeCP2 fusion proteins in the cell extracts.

FIG. 5.

The CRID/TRD region of MeCP2 is required to relocalize LANA to chromocenters. Cells were cotransfected with plasmids expressing GFP-LANA_FL and full-length or truncated versions of DsRed-MeCP2. Single cells displaying the predominant pattern for each transfection experiment are shown.

FIG. 6.

LANA binds to the multiple domains within MeCP2. In vitro binding analysis using recombinant GST-LANA_C purified from bacteria and ³⁵S-labeled full-length or truncated MeCP2 (shown schematically below) synthesized by in vitro translation. Binding reactions were performed with 100 mM NaCl, and mixtures were washed in 250 mM NaCl. ³⁵S-labeled LANA_C (lanes 6, 12, and 18) served as a positive control for binding.

FIG. 7.

LANA and MeCP2 cooperate to regulate transcription. (A) LANA bound to reiterated copies of the KSHV TR sequence can repress a constitutively active promoter located on the same plasmid. The schematic shows MeCP2 truncations, indicating the three domains (MBD, CRID, and TRD) required for association with LANA. (B) HeLa cells were cotransfected with the pGL3p-7xTR reporter and expression plasmids encoding derivatives of LANA and/or MeCP2. Luciferase activity was measured after 24 h, and the degree of repression (n-fold) was calculated relative to expression in cells with the reporter alone. Values represent the means and standard deviations of results from three independent assays. The dotted line indicates the level of repression by LANA_NC alone. Note that expression of MeCP2 in the absence of LANA results in modest stimulation. (C) Results of a repression assay comparing the LANA_FL wild type (WT) and the CBM deletion mutant (Δ22) in the presence or absence of MeCP2. (D) Stimulation of the human E2F1 promoter by LANA and MeCP2 is sensitive to mutations in the LANA CBM. Cells were cotransfected with the pE2F1-luc reporter and plasmids expressing WT and mutant versions of LANA_NC in the presence or absence of MeCP2_FL. The dotted line corresponds to reporter activity in the presence of LANA_NC alone.

FIG. 8.

Schematics of possible multivalent interactions among LANA, core histones, and MeCP2. (i) Evidence suggests that LANA may be folded such that LANA_N, which includes the CBM, is brought into proximity to LANA_C. For simplicity, only one molecule of LANA is shown; however, LANA most likely exists as a stable dimer. (ii) Engagement of the CBM with a nucleosome may alter the conformation, making the C-terminal domain more accessible to secondary interactions. (iii) Interaction of LANA_C with MeCP2 forms a relatively stable complex mediated by contacts between the CBM and the nucleosome core (a) and DNA and the MeCP2 MBD (b). As a consequence of these multivalent interactions, LANA accumulates with MeCP2 at the chromocenters. (iv) In the absence of the CBM, the LANA protein (LANA_FLΔ22) shows reduced retention at the chromocenters. This may occur because the deletion alters the accessibility of the LANA_C domain or increases the rate at which LANA exchanges with other interaction partners located in more euchromatic regions of the genome.